US20060195497A1 - Method, apparatus and program storage device that provides a shift process with saturation for digital signal processor operations - Google Patents
Method, apparatus and program storage device that provides a shift process with saturation for digital signal processor operations Download PDFInfo
- Publication number
- US20060195497A1 US20060195497A1 US11/068,206 US6820605A US2006195497A1 US 20060195497 A1 US20060195497 A1 US 20060195497A1 US 6820605 A US6820605 A US 6820605A US 2006195497 A1 US2006195497 A1 US 2006195497A1
- Authority
- US
- United States
- Prior art keywords
- saturation
- memory location
- shifter
- data
- output
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title abstract description 24
- 230000008569 process Effects 0.000 title abstract description 14
- 238000001514 detection method Methods 0.000 claims abstract description 47
- 230000004044 response Effects 0.000 claims description 17
- 238000013500 data storage Methods 0.000 claims description 5
- 229920006395 saturated elastomer Polymers 0.000 abstract description 16
- BMQYVXCPAOLZOK-NJGYIYPDSA-N D-monapterin Chemical compound C1=C([C@H](O)[C@@H](O)CO)N=C2C(=O)NC(N)=NC2=N1 BMQYVXCPAOLZOK-NJGYIYPDSA-N 0.000 description 12
- 230000000295 complement effect Effects 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 238000007792 addition Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000009738 saturating Methods 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11B—INFORMATION STORAGE BASED ON RELATIVE MOVEMENT BETWEEN RECORD CARRIER AND TRANSDUCER
- G11B20/00—Signal processing not specific to the method of recording or reproducing; Circuits therefor
- G11B20/10—Digital recording or reproducing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F5/00—Methods or arrangements for data conversion without changing the order or content of the data handled
- G06F5/01—Methods or arrangements for data conversion without changing the order or content of the data handled for shifting, e.g. justifying, scaling, normalising
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F7/00—Methods or arrangements for processing data by operating upon the order or content of the data handled
- G06F7/38—Methods or arrangements for performing computations using exclusively denominational number representation, e.g. using binary, ternary, decimal representation
- G06F7/48—Methods or arrangements for performing computations using exclusively denominational number representation, e.g. using binary, ternary, decimal representation using non-contact-making devices, e.g. tube, solid state device; using unspecified devices
- G06F7/499—Denomination or exception handling, e.g. rounding or overflow
- G06F7/49905—Exception handling
- G06F7/4991—Overflow or underflow
- G06F7/49921—Saturation, i.e. clipping the result to a minimum or maximum value
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11B—INFORMATION STORAGE BASED ON RELATIVE MOVEMENT BETWEEN RECORD CARRIER AND TRANSDUCER
- G11B2220/00—Record carriers by type
- G11B2220/20—Disc-shaped record carriers
- G11B2220/25—Disc-shaped record carriers characterised in that the disc is based on a specific recording technology
- G11B2220/2508—Magnetic discs
- G11B2220/2516—Hard disks
Definitions
- This invention relates in general to digital signal processing, and more particularly to a method, apparatus and program storage device that provides a shift process with saturation for digital signal processor operations.
- the central processing unit includes circuitry to receive and decode instructions and circuitry to process data in accordance with the decoded instructions.
- the circuitry to process the data typically includes an arithmetic logic unit (ALU).
- ALU arithmetic logic unit
- the arithmetic logic unit performs arithmetic functions such as add, subtract, multiply, divide, shift data, etc. and performs logic functions such as AND, OR, NAND, NOR, exclusive OR, etc. More specifically, the ALU consists of two kinds of operations: arithmetic operation such as addition/subtraction and logical operation.
- Addition and subtraction are straightforward and unexceptional.
- the result may be an n+1-bit binary due to the carry-out.
- data is typically represented as fixed-width number.
- the sum of 0110 and 1110 will be 0100 in a 4-bit processor and an overflow is used to point out the result is erroneous.
- Subtraction behaves similar. The same problem occurs in two's complement number system, under different condition. In the twos complement representation, negative numbers are represented as the result of a subtraction of the magnitude of the number from zero.
- a negative 2 is represented as the result of 0000 minus 0010, which is 1110 (the borrow, or carry, produced by the subtraction is ignored).
- the twos complement representation has the advantageous characteristic that additions and subtractions of twos complement numbers can be effected without regard for whether the numbers being added or subtracted are positive or negative.
- the overflow occurs when adding two positive numbers and the sum is negative, or vice versa.
- adding or subtracting two 32-bit numbers can yield a result that needs 33-bits to be fully expressed.
- the lack of a 33rd bit means that the overflow occurs and the sign bit is being set with the value of the result instead of the proper sign of the result.
- the arithmetic logic unit includes an arithmetic shifter.
- An arithmetic shifter is used to store a bit-vector of some specified length. It has two control signals: shift-left and shift-right. When shift-left is asserted, the bits in the vector are shifted one bit to the left, with a 0 shifted into the rightmost bit. When shift-right is asserted, the bits in the vector are shifted one bit to the right, with a copy of the most significant bit shifted into the leftmost bit.
- An arithmetic shifter is typically implemented using a shift register.
- a shift register is a group of registers set up in a linear fashion that have their inputs and outputs connected together in such a way that the data is shifted down the line when the circuit is activated. Shifting a word right or left (which is equivalent to multiplying or dividing by a power of 2) is used in multiplication and division and also to align data on byte or word boundaries.
- Arithmetic overflow is the condition that occurs when a calculation produces a result that is greater than a given register or storage location can store or represent.
- a digital processor may use a self-saturating number representation in which any arithmetic operation which generates an overflow will automatically set the output to a value having the proper sign and a magnitude equal to the maximum value which can be represented by the digital data word.
- ALU saturation logic may be used to prevent a result from overflowing by keeping the result at a maximum (or minimum) value. Thus, when arithmetic operations produce values too large or too small for registers, the largest or smallest value that can be represented is substituted instead. More specifically:
- the present invention discloses a method, apparatus and program storage device that provides a shift process with saturation for digital signal processor operations.
- a saturation detection circuit is provided with an arithmetic shifter and a final decision multiplexor that receives the output from the arithmetic shifter and the saturated value from a saturation circuit.
- the term multiplexor is intended to include any type of device, e.g., switch, logic device, programming code, etc., that connects one signal selected from several inputs to a single output.
- the final decision multiplexor selects the saturate minimum or the saturate maximum depending on whether the MSB of the data in equals one or zero, respectively.
- An arithmetic shifter with saturation detection in accordance with the principles of the present invention includes an arithmetic shifter for receiving input data and a shift amount indication, a saturation detector, coupled to the arithmetic shifter, for detecting when saturation occurs and generating a saturation signal in response, a saturation value generator for generating a saturation value and an output selector for selecting whether to provide as an output shifter the output data or the saturation value in response to the saturation signal.
- a processor unit in another embodiment, includes memory for storing data and instructions therein and a processor for performing arithmetic shift operations, the processor being configured to receive input data and a shift amount indication, the processor further being configured for detecting when saturation occurs, generating a saturation value for output when saturation occurs and for providing as an output either the input data shifted by the shift amount or, in response to saturation being detected, the saturation value.
- a storage control device in another embodiment, includes a storage controller for processing read and write signals and a processor unit, coupled to the storage controller, the processor unit performing operations in support of storage operations, wherein the processor unit is configured to receive a shift instruction from the storage controller, the shift instruction allowing input data to be shifted N bits in a single cycle, the processor further being configured for detecting when saturation results from the shift operation, generating a saturation value for output when saturation occurs and for providing as an output either the input data shifted by N bits or, in response to saturation being detected, the saturation value.
- a data storage system in another embodiment, includes a translatable recording medium for storing data thereon, a motor for translating the recording medium, a transducer disposed proximate to the recording medium for reading and writing data on the recording medium, an actuator, coupled to the transducer, for moving the transducer relative to the recording medium and a storage control device for controlling operations of the data storage system, the storage control device further including a storage controller for processing read and write signals and a processor unit, coupled to the storage controller, the processor unit performing operations in support of storage operations, wherein the processor unit is configured to receive a shift instruction from the storage controller, the shift instruction allowing input data to be shifted N bits in a single cycle, the processor further being configured for detecting when saturation results from the shift operation, generating a saturation value for output when saturation occurs and for providing as an output either the input data shifted by N bits or, in response to saturation being detected, the saturation value.
- an arithmetic shifter with saturation detection includes means for receiving input data and a shift amount indication, means, coupled to the means for receiving input data and a shift amount indication, for detecting when saturation occurs and generating a saturation signal in response, means for generating a saturation value and means for selecting to provide as an output shifter output data or the saturation value in response to the saturation signal.
- a processor unit in another embodiment, includes means for storing data and instructions therein and means for performing arithmetic shift operations, receiving input data and a shift amount indication, detecting when saturation occurs, generating a saturation value for output when saturation occurs and providing as an output either the input data shifted by the shift amount or, in response to saturation being detected, the saturation value.
- FIG. 1 illustrates an arithmetic shifter and saturation detection circuit according to an embodiment of the present invention
- FIG. 2 illustrates a block diagram of a digital signal processing circuit for implementing an arithmetic shifter and saturation detection circuit according to an embodiment of the present invention
- FIG. 3 is a flow chart of the method for providing a shift process with saturation for digital signal processor operations according to an embodiment of the present invention.
- FIG. 4 shows a schematic block diagram showing a hard disk storage system according to one embodiment of the present invention.
- the present invention provides a method, apparatus and program storage device that provides a shift process with saturation for digital signal processor operations.
- a single instruction is generated for shifting an operand to either maximum or the minimum value depending on the bit of data input when saturation occurs.
- a saturation detection circuit is provided with an arithmetic shifter and a final decision multiplexor that receives the output from the arithmetic shifter and the saturated value from a saturation circuit. When saturation is detected, the final decision multiplexor selects the saturate minimum or the saturate maximum depending on whether the MSB of the data in equals one or zero, respectively.
- FIG. 1 illustrates an arithmetic shifter and saturation detection circuit 100 according to an embodiment of the present invention.
- data 110 and shift amount 112 values are provided to an arithmetic shifter 120 .
- the arithmetic shifter 120 provides an output to maximum shifter 122 and to minimum 124 shifter.
- An underflow 130 and overflow 132 detection circuit are coupled to the maximum shifter 122 and to the minimum shifter 124 respectively.
- the arithmetic shifter 120 shifts data thru maximum shifter 122 and minimum shifter 124 , the bits leaving the arithmetic shifter 120 are filling the maximum shifter 122 and the minimum shifter 124 .
- the arithmetic shift amount 112 is “n”
- the arithmetic shifter 120 can perform from 0 to (2n-1) bit left shifts in a cycle.
- a saturation decision multiplexor 140 receives input from the underflow 130 and overflow 132 detection circuits.
- the term multiplexor is intended to include any type of device, e.g., switch, logic device, programming code, etc., that connects one signal selected from several inputs to a single output.
- the most significant bit of data input 142 is used to control the saturation decision multiplexor 140 .
- a final decision multiplexor 150 receives the output 194 from the arithmetic shifter 120 , and the saturated value 160 from the output multiplexor 170 .
- the saturation decision multiplexor 140 provides an indicator of saturation to an OR gate 180 .
- the most significant bit of the data 182 and the most significant bit of arithmetic shifter 184 are exclusively ORed (XOR) 186 . This result is ORed 180 with the indication of saturation from the saturation decision multiplexor 140 . If either is high, the signal “saturation detected” 190 is high.
- the arithmetic shifter and saturation detection circuit 100 operates as follows. When the multiple bits are shifted left in a cycle, an arithmetic shifter 120 alone is not able to determine by the arithmetic shifter output if the arithmetic shifter 120 has had a situation of overflow and underflow during the shift operation. Basically, there are two situations when the saturation can be applied. One is that the sign of the arithmetic shifter 120 is different from the sign of the data in 110 . The other is when the bits 194 leaving the arithmetic shifter 120 are not the same, i.e., all bits are not either zeros or ones.
- the arithmetic shifter and saturation detection circuit 100 detects the saturation cases and modifies the shifter output to either the maximum or the minimum value depending on the sign bit of data input when the saturation has occurred.
- the arithmetic shifter and saturation detection circuit 100 shifts an operand to the left, i.e., 2 x *operand.
- the result would be saturated to (2 n -1) or -2 n , which ever is appropriate, for the given sign. Operations would then be the same in the sense that negative values would remain negative and positive values would remain positive.
- the content of 15 bit maximum shifter 122 is 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, D 31 , D 30 , D 29 .
- the ones detection circuit 130 checks if all bits of the shifter 122 are one.
- the ones detection circuit 130 may be implemented by 15 bit “AND” gate. When the output of ones detection circuit 130 is low, the arithmetic shifter 120 may have been underflowed.
- the content of 15 bit minimum shifter 124 is 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, D 31 , D 30 , D 29 .
- the zeros detection circuit 132 checks if all bits are of the shifter 124 are zero.
- the zeros detection circuit 132 may be implemented by 15 bit “NOR” gate. When the output of zeros detection circuit 132 is low, the arithmetic shifter 120 may have been overflowed.
- the saturation decision multiplexor 140 is selected by the most significant bit (MSB) 142 of data input, D( 31 ). If the MSB 142 , D( 31 ), is one, saturation decision multiplexor 140 selects the output of the ones detection circuit 130 ; otherwise, the saturation decision multiplexor 140 selects the output of the zeros detection circuit 132 . When the output signal of the saturation decision multiplexor 140 is low, saturation has occurred.
- the maximum saturated value (overflow) and the minimum saturated value (underflow) are defined to be “7FFF FFFF” and “8000 0000/1” (could be 8000 0001 for symmetry with positive maximum value) respectively.
- the output of the output multiplexor 170 is “8000 0000/1” when the MSB 142 , D( 31 ), is one; otherwise the output of the output multiplexor 170 is “7FFF FFFF”.
- the final decision multiplexor 150 selects either the output 194 of the arithmetic shifter 120 or the saturated value 160 from the output multiplexor 170 .
- the shift output is modified to the saturated value 160 , otherwise the output 194 of the arithmetic shifter 120 .
- Saturation is determined using the output of the saturation decision multiplexor 140 and the most significant bit, S( 31 ) 184 , of the arithmetic shifter output, S( 31 : 0 ) 194 , and the most significant bit, D( 31 ) 182 , of Data in, ( 31 : 0 ) 110 .
- the final decision multiplexor 150 selects the saturate minimum or the saturate maximum 160 depending on the MSB 142 , i.e., D( 31 ) equals one or zero, respectively.
- D( 31 ) equals one or zero, respectively.
- the saturation detection is based on the MSB of the 10 arithmetic shifter output, S 31 184 , and the MSB of the data, D 31 182 , being different, wherein the MSB of the data, D 31 142 , being zero selects the output of the output multiplexor to be 7FFF FFFF.
- the saturation detection is based on the output of the zero detection circuit being zero (i.e., the minimum shift register does not contain all zeroes).
- the saturation detection is based on the MSB of the arithmetic shifter output, S 31 184 , and the MSB of the data, D 31 182 , being different, wherein the MSB of the data, D 31 142 , being one selects the output of the output multiplexor to be 8000 0000.
- the saturation detection is based on the output of the ones detection circuit being zero (i.e., the maximum shift register does not contain all ones).
- FIG. 2 illustrates a block diagram of a digital signal processing circuit 200 for implementing an arithmetic shifter and saturation detection circuit according to an embodiment of the present invention.
- FIG. 2 shows a control unit 210 , a data unit 220 and an address unit 230 .
- the control unit 210 directs the operation of the digital signal processor based on an instruction set (ISA) optimised for the task of rapid signal processing.
- ISA instruction set
- the signal processing is divided between the control unit 210 that directs program flow and one or more execution units that perform operations on data. Almost always, a collection of registers/memory 240 is included to hold operands and intermediate results.
- One of the execution units is the address unit 230 .
- the address unit 230 directs the operand fetch for all variables which are defined and used by the executing instructions or program.
- Another execution unit is the data unit 220 , which includes at least one arithmetic logic unit 222 that accepts as inputs the data to be operated on and a code from the control unit 210 indicating what operation to perform.
- the arithmetic logic unit 222 provides the result of the computation and may indicate cases such as carry-in or carry-out, overflow, underflow and/or other statuses.
- FIG. 3 is a flow chart 300 of the method for providing a shift process with saturation for digital signal processor operations according to an embodiment of the present invention.
- data and shift amount values are provided to an arithmetic shifter 310 .
- the arithmetic shifter provides an output to maximum shifter and to minimum shifter 312 .
- An underflow and overflow detection circuit detect underflow and overflow based on the content of the minimum shifter and the maximum shifter 314 .
- a saturation decision multiplexor receives input from the underflow and overflow detection circuits 316 .
- a final decision multiplexor receives the output from the arithmetic shifter and the saturated value from the output multiplexor 318 .
- Whether saturation occurs is determined by the final decision multiplexor using the output of the saturation decision multiplexor, the most significant bit of the arithmetic shifter output, and the most significant bit of data in 320 . Saturation is detected if the most significant bit of the arithmetic shifter output and the most significant bit of data in are different or if the saturation decision multiplexor so indicates.
- the final decision multiplexor selects the saturate minimum or the saturate maximum depending on whether the MSB equals one or zero, respectively 330 .
- the final decision multiplexor selects the output from the arithmetic shifter 340 .
- a digital signal processor as illustrated above with respect to FIGS. 1-3 may be used in a wide variety of products such as field programmable gate arrays, personal digital assistants (PDAs), cellular phones, digital scanners, digital cameras, personal computers and storage controllers.
- FIG. 4 shows one example wherein a magnetic disk drive includes a hard drive controller and microprocessor unit.
- the hard drive controller, microprocessor and memory e.g., SRAM, may be shown separately, such components could easily be combined into one component, e.g., the hard drive controller.
- FIG. 4 shows a schematic block diagram showing a hard disk storage system 400 according to one embodiment of the present invention.
- the hard disk storage system 400 is connected to a host computer 432 .
- the hard disk storage system 400 responds to the write request by the host computer 432 and records the recording data from the host computer 432 on a magnetic disk 410 , which serves as a recording medium.
- the hard disk storage system 400 further responds to the read request from the host computer 432 , reads the data recorded on the magnetic disk 410 , and sends the data to the host computer 432 .
- the hard disk storage system 400 includes the magnetic disk 410 , first and second motors 412 , 416 , a head device 414 , a signal processing circuit 420 , a servo circuit 430 , a microprocessor (MPU) 440 , a memory (RAM) 450 , a hard drive controller (HDC) 460 , and an interface circuit 470 .
- the circuits 420 - 470 are connected to one another by a bus 480 .
- the magnetic disk 410 is rotated by the first motor 412 at a constant rotating speed.
- the second motor 416 controls the head device 414 so that it moves in the radial direction with respect to the magnetic disk 410 .
- the head device 414 reads the data recorded on the magnetic disk 410 and sends a read signal, RD, to the signal processing circuit 420 .
- the signal processing circuit 420 samples the read signal, RD, in synchronism with a clock signal and generates a digital read signal.
- the signal processing circuit 420 carries out a decoding process on the digital read signal and outputs the decoded data signal.
- the servo circuit 430 controls the first motor 412 and rotates the magnetic disk 410 at a constant speed.
- the servo circuit 430 further receives the decoded data signal from the signal processing circuit 420 via the bus 480 and controls the second motor 416 based on the servo data included in the digital read signal so that the head device 414 is on track at the target position.
- the MPU 440 analyzes the write/read processing command sent from the host computer 432 in accordance with the program data stored in the RAM 450 and sends a control signal to the HDC 460 via the bus 480 .
- the HDC 460 controls the signal processing circuit 420 and the servo circuit 430 in accordance with the control signal from the MPU 440 .
- the HDC 460 further receives a data signal from the signal processing circuit 420 via the bus 480 .
- the HDC 460 processes date, e.g., performs an error correcting code (ECC) process on the data signal.
- ECC error correcting code
- the interface circuit 470 converts the data from the HDC 460 to a predetermined communication mode and sends the converted data to the host computer 432 .
- the MPU 440 includes an arithmetic logic unit including an arithmetic shifter and saturation detection circuit as illustrated in FIG. 1 for shifting an operand to either maximum or the minimum value depending on the bit of data input when saturation occurs.
- the HDC 460 sends a single shift instruction to the MPU 440 .
- the MPU 440 may be configured so that the shift instruction may be run by firmware of the HDC 460 .
- the MPU 440 returns a shifted operand or a minimum/maximum saturation value.
- the MPU 440 could include a standalone processor or an embedded processor, e.g., the MPU 440 could be embedded in the HDC 460 .
- the MPU 440 could be part of a system on a chip (SOC).
- the MPU 440 could be an ASIC, which would be hardware circuits that perform the function of the processor operating pursuant to memory 450 .
- memory 450 may be used but is not required, as the ASIC is designed to perform any assigned functions.
- memory 450 could be either volatile or non-volatile memory.
- the MPU 440 controls the operation of the voice coil motor 416 and spindle motor 412 via the servo unit 430 .
Abstract
Description
- 1. Field of the Invention
- This invention relates in general to digital signal processing, and more particularly to a method, apparatus and program storage device that provides a shift process with saturation for digital signal processor operations.
- 2. Description of Related Art
- Techniques are known in the domain of integrated circuits using calculation units for verifying if the format allowed for the result of an arithmetic operation carried out by an AU (Arithmetic Unit) on two operands and an input carry digit does not exceed a given format. This technique has applications in most calculation units, such as, calculation units included in programmable circuits, such as, a digital signal processor (DSP) or a microcontroller.
- To process the applications, the central processing unit includes circuitry to receive and decode instructions and circuitry to process data in accordance with the decoded instructions. The circuitry to process the data typically includes an arithmetic logic unit (ALU). The arithmetic logic unit performs arithmetic functions such as add, subtract, multiply, divide, shift data, etc. and performs logic functions such as AND, OR, NAND, NOR, exclusive OR, etc. More specifically, the ALU consists of two kinds of operations: arithmetic operation such as addition/subtraction and logical operation.
- Addition and subtraction are straightforward and unexceptional. When adding two i-bit numbers, the result may be an n+1-bit binary due to the carry-out. In computer hardware, data is typically represented as fixed-width number. Thus, the sum of 0110 and 1110 will be 0100 in a 4-bit processor and an overflow is used to point out the result is erroneous. Subtraction behaves similar. The same problem occurs in two's complement number system, under different condition. In the twos complement representation, negative numbers are represented as the result of a subtraction of the magnitude of the number from zero. That is, in a four-bit system, for example, a negative 2 is represented as the result of 0000 minus 0010, which is 1110 (the borrow, or carry, produced by the subtraction is ignored). The twos complement representation has the advantageous characteristic that additions and subtractions of twos complement numbers can be effected without regard for whether the numbers being added or subtracted are positive or negative. In the two's complement number system, the overflow occurs when adding two positive numbers and the sum is negative, or vice versa. Clearly, adding or subtracting two 32-bit numbers can yield a result that needs 33-bits to be fully expressed. The lack of a 33rd bit means that the overflow occurs and the sign bit is being set with the value of the result instead of the proper sign of the result.
- Multiplication of signed numbers requires special care. For example, multiplying numbers with the same sign produces a positive product, but multiplying a positive number by a negative number yields a negative product. Multiplication may be performed using a sequence of shifts and additions. To achieve the data shift function, the arithmetic logic unit includes an arithmetic shifter. An arithmetic shifter is used to store a bit-vector of some specified length. It has two control signals: shift-left and shift-right. When shift-left is asserted, the bits in the vector are shifted one bit to the left, with a 0 shifted into the rightmost bit. When shift-right is asserted, the bits in the vector are shifted one bit to the right, with a copy of the most significant bit shifted into the leftmost bit.
- An arithmetic shifter is typically implemented using a shift register. A shift register is a group of registers set up in a linear fashion that have their inputs and outputs connected together in such a way that the data is shifted down the line when the circuit is activated. Shifting a word right or left (which is equivalent to multiplying or dividing by a power of 2) is used in multiplication and division and also to align data on byte or word boundaries.
- ALU results are saturated upon overflow. Arithmetic overflow is the condition that occurs when a calculation produces a result that is greater than a given register or storage location can store or represent. A digital processor may use a self-saturating number representation in which any arithmetic operation which generates an overflow will automatically set the output to a value having the proper sign and a magnitude equal to the maximum value which can be represented by the digital data word. For example, ALU saturation logic may be used to prevent a result from overflowing by keeping the result at a maximum (or minimum) value. Thus, when arithmetic operations produce values too large or too small for registers, the largest or smallest value that can be represented is substituted instead. More specifically:
-
- if ((result=a±b)>(2n-1))
- result=2n-1;
- else if ((a±b)<=−2n)
- result=−2n;
However, there is typically no solution for saturating the result of 2x*data. Furthermore, solutions that are used are very expensive in terms of cycles.
- result=−2n;
- if ((result=a±b)>(2n-1))
- It can be seen then that there is a need for a method, apparatus and program storage device that provides a shift process with saturation for digital signal processor operations.
- To overcome the limitations in the prior art described above, and to overcome other limitations that will become apparent upon reading and understanding the present specification, the present invention discloses a method, apparatus and program storage device that provides a shift process with saturation for digital signal processor operations.
- The present invention solves the above-described problems by providing an instruction for shifting an operand to either maximum or the minimum value depending on the most significant bit of data input when saturation occurs. A saturation detection circuit is provided with an arithmetic shifter and a final decision multiplexor that receives the output from the arithmetic shifter and the saturated value from a saturation circuit. Herein, the term multiplexor is intended to include any type of device, e.g., switch, logic device, programming code, etc., that connects one signal selected from several inputs to a single output. When saturation is detected, the final decision multiplexor selects the saturate minimum or the saturate maximum depending on whether the MSB of the data in equals one or zero, respectively.
- An arithmetic shifter with saturation detection in accordance with the principles of the present invention includes an arithmetic shifter for receiving input data and a shift amount indication, a saturation detector, coupled to the arithmetic shifter, for detecting when saturation occurs and generating a saturation signal in response, a saturation value generator for generating a saturation value and an output selector for selecting whether to provide as an output shifter the output data or the saturation value in response to the saturation signal.
- In another embodiment of the present invention, a processor unit is disclosed. The processor unit includes memory for storing data and instructions therein and a processor for performing arithmetic shift operations, the processor being configured to receive input data and a shift amount indication, the processor further being configured for detecting when saturation occurs, generating a saturation value for output when saturation occurs and for providing as an output either the input data shifted by the shift amount or, in response to saturation being detected, the saturation value.
- In another embodiment of the present invention, a storage control device is disclosed. The storage control device includes a storage controller for processing read and write signals and a processor unit, coupled to the storage controller, the processor unit performing operations in support of storage operations, wherein the processor unit is configured to receive a shift instruction from the storage controller, the shift instruction allowing input data to be shifted N bits in a single cycle, the processor further being configured for detecting when saturation results from the shift operation, generating a saturation value for output when saturation occurs and for providing as an output either the input data shifted by N bits or, in response to saturation being detected, the saturation value.
- In another embodiment of the present invention, a data storage system is disclosed. The data storage system includes a translatable recording medium for storing data thereon, a motor for translating the recording medium, a transducer disposed proximate to the recording medium for reading and writing data on the recording medium, an actuator, coupled to the transducer, for moving the transducer relative to the recording medium and a storage control device for controlling operations of the data storage system, the storage control device further including a storage controller for processing read and write signals and a processor unit, coupled to the storage controller, the processor unit performing operations in support of storage operations, wherein the processor unit is configured to receive a shift instruction from the storage controller, the shift instruction allowing input data to be shifted N bits in a single cycle, the processor further being configured for detecting when saturation results from the shift operation, generating a saturation value for output when saturation occurs and for providing as an output either the input data shifted by N bits or, in response to saturation being detected, the saturation value.
- In another embodiment of the present invention, an arithmetic shifter with saturation detection is disclosed. The arithmetic shifter with saturation detection includes means for receiving input data and a shift amount indication, means, coupled to the means for receiving input data and a shift amount indication, for detecting when saturation occurs and generating a saturation signal in response, means for generating a saturation value and means for selecting to provide as an output shifter output data or the saturation value in response to the saturation signal.
- In another embodiment of the present invention, a processor unit is disclosed. The processor unit includes means for storing data and instructions therein and means for performing arithmetic shift operations, receiving input data and a shift amount indication, detecting when saturation occurs, generating a saturation value for output when saturation occurs and providing as an output either the input data shifted by the shift amount or, in response to saturation being detected, the saturation value.
- These and various other advantages and features of novelty which characterize the invention are pointed out with particularity in the claims annexed hereto and form a part hereof. However, for a better understanding of the invention, its advantages, and the objects obtained by its use, reference should be made to the drawings which form a further part hereof, and to accompanying descriptive matter, in which there are illustrated and described specific examples of an apparatus in accordance with the invention.
- Referring now to the drawings in which like reference numbers represent corresponding parts throughout:
-
FIG. 1 illustrates an arithmetic shifter and saturation detection circuit according to an embodiment of the present invention; -
FIG. 2 illustrates a block diagram of a digital signal processing circuit for implementing an arithmetic shifter and saturation detection circuit according to an embodiment of the present invention; -
FIG. 3 is a flow chart of the method for providing a shift process with saturation for digital signal processor operations according to an embodiment of the present invention; and -
FIG. 4 shows a schematic block diagram showing a hard disk storage system according to one embodiment of the present invention. - In the following description of the embodiments, reference is made to the accompanying drawings that form a part hereof, and in which is shown by way of illustration the specific embodiments in which the invention may be practiced. It is to be understood that other embodiments may be utilized because structural changes may be made without departing from the scope of the present invention.
- The present invention provides a method, apparatus and program storage device that provides a shift process with saturation for digital signal processor operations. A single instruction is generated for shifting an operand to either maximum or the minimum value depending on the bit of data input when saturation occurs. A saturation detection circuit is provided with an arithmetic shifter and a final decision multiplexor that receives the output from the arithmetic shifter and the saturated value from a saturation circuit. When saturation is detected, the final decision multiplexor selects the saturate minimum or the saturate maximum depending on whether the MSB of the data in equals one or zero, respectively.
-
FIG. 1 illustrates an arithmetic shifter andsaturation detection circuit 100 according to an embodiment of the present invention. InFIG. 1 ,data 110 andshift amount 112 values are provided to anarithmetic shifter 120. At the beginning of a cycle, thearithmetic shifter 120 provides an output tomaximum shifter 122 and tominimum 124 shifter. Anunderflow 130 andoverflow 132 detection circuit are coupled to themaximum shifter 122 and to theminimum shifter 124 respectively. When thearithmetic shifter 120 shifts data thrumaximum shifter 122 andminimum shifter 124, the bits leaving thearithmetic shifter 120 are filling themaximum shifter 122 and theminimum shifter 124. If thearithmetic shift amount 112 is “n”, thearithmetic shifter 120 can perform from 0 to (2n-1) bit left shifts in a cycle. - A
saturation decision multiplexor 140 receives input from theunderflow 130 andoverflow 132 detection circuits. Herein, the term multiplexor is intended to include any type of device, e.g., switch, logic device, programming code, etc., that connects one signal selected from several inputs to a single output. The most significant bit ofdata input 142 is used to control thesaturation decision multiplexor 140. Afinal decision multiplexor 150 receives theoutput 194 from thearithmetic shifter 120, and the saturatedvalue 160 from theoutput multiplexor 170. Thesaturation decision multiplexor 140 provides an indicator of saturation to anOR gate 180. The most significant bit of thedata 182 and the most significant bit ofarithmetic shifter 184 are exclusively ORed (XOR) 186. This result isORed 180 with the indication of saturation from thesaturation decision multiplexor 140. If either is high, the signal “saturation detected” 190 is high. - The arithmetic shifter and
saturation detection circuit 100 operates as follows. When the multiple bits are shifted left in a cycle, anarithmetic shifter 120 alone is not able to determine by the arithmetic shifter output if thearithmetic shifter 120 has had a situation of overflow and underflow during the shift operation. Basically, there are two situations when the saturation can be applied. One is that the sign of thearithmetic shifter 120 is different from the sign of the data in 110. The other is when thebits 194 leaving thearithmetic shifter 120 are not the same, i.e., all bits are not either zeros or ones. - The arithmetic shifter and
saturation detection circuit 100 detects the saturation cases and modifies the shifter output to either the maximum or the minimum value depending on the sign bit of data input when the saturation has occurred. The arithmetic shifter andsaturation detection circuit 100 shifts an operand to the left, i.e., 2 x*operand. The result would be saturated to (2n-1) or -2n, which ever is appropriate, for the given sign. Operations would then be the same in the sense that negative values would remain negative and positive values would remain positive. - Assume
data 110 provided to thearithmetic shifter 120 was D31, D30, D29, . . . , D0. At the beginning of the cycle, themaximum shifter 122 is initially filled with ones, and theminimum shifter 124 is initially filled with zeroes. If, for example, the shift amount is 3, three bits leave the arithmetic shifter, D(31:29), and fill the maximum 122 and minimum 124 shifter with D(31:29) from the right. The output of thearithmetic shifter 120 is D28, D27, D26, . . . , D0, “000”. The content of 15 bitmaximum shifter 122 is 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, D31, D30, D29. Theones detection circuit 130 checks if all bits of theshifter 122 are one. Theones detection circuit 130 may be implemented by 15 bit “AND” gate. When the output ofones detection circuit 130 is low, thearithmetic shifter 120 may have been underflowed. - The content of 15 bit
minimum shifter 124 is 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, D31, D30, D29. Thezeros detection circuit 132 checks if all bits are of theshifter 124 are zero. Thezeros detection circuit 132 may be implemented by 15 bit “NOR” gate. When the output ofzeros detection circuit 132 is low, thearithmetic shifter 120 may have been overflowed. - The
saturation decision multiplexor 140 is selected by the most significant bit (MSB) 142 of data input, D(31). If theMSB 142, D(31), is one,saturation decision multiplexor 140 selects the output of theones detection circuit 130; otherwise, thesaturation decision multiplexor 140 selects the output of thezeros detection circuit 132. When the output signal of thesaturation decision multiplexor 140 is low, saturation has occurred. The maximum saturated value (overflow) and the minimum saturated value (underflow) are defined to be “7FFF FFFF” and “8000 0000/1” (could be 8000 0001 for symmetry with positive maximum value) respectively. The output of theoutput multiplexor 170 is “8000 0000/1” when theMSB 142, D(31), is one; otherwise the output of theoutput multiplexor 170 is “7FFF FFFF”. - The
final decision multiplexor 150 selects either theoutput 194 of thearithmetic shifter 120 or the saturatedvalue 160 from theoutput multiplexor 170. When the signal “saturation detected” 190 is high, the shift output is modified to the saturatedvalue 160, otherwise theoutput 194 of thearithmetic shifter 120. Saturation is determined using the output of thesaturation decision multiplexor 140 and the most significant bit, S(31) 184, of the arithmetic shifter output, S(31:0) 194, and the most significant bit, D(31) 182, of Data in, (31:0)110. First, when the data in 110 is shifted byn 112, the result is in thearithmetic shifter 120. When the most significant bit, S(31) 184, of theoutput 194 of thearithmetic shifter 120, and the most significant bit, D(31) 182, of Data in, (31:0) 110, are different, the saturation has occurred. Second, when the output ofsaturation decision multiplexor 140 is low, the saturation has occurred. In either case, the signal “saturation detected” 190 is set to high. When the signal “saturation detected” 190 is high, saturation has occurred. Thefinal decision multiplexor 150 selects the saturate minimum or thesaturate maximum 160 depending on theMSB 142, i.e., D(31) equals one or zero, respectively.TABLE 1 MSB Arith MSB Final Data in Data Shift shifter Shifter Zeros Ones Saturate Saturate Output (31:0) D(31) amt(n:0) (31:0) S(31) detect detect detected value (31:0) Normal Case 0x07FF 0 3 0x3FFF 0 1 0 0 0x7FFF 0x3FFF FFFF FFF8 FFFF FFF8 0xF7FF 1 3 0xBFFF 1 0 1 0 0x8000 0xBFFF FFFF FFF8 0000 FFF8 Saturated Case 0x08FF 0 4 0x8FFF 1 1 0 1 0x7FFF 0x7FFF FFFF FFF0 FFFF FFFF 0x08FF 0 5 0x1FFF 0 0 0 1 0x7FFF 0x7FFF FFFF FFE0 FFFF FFFF 0xF7FF 1 4 0x7FFF 0 0 1 1 0x8000 0x8000 FFFF FFF0 0000 0000 0xF7FF 1 5 0xFFFF 1 0 0 1 0x8000 0x8000 FFFF FFE0 0000 0000 - In the first saturated case, the saturation detection is based on the MSB of the 10 arithmetic shifter output,
S31 184, and the MSB of the data,D31 182, being different, wherein the MSB of the data,D31 142, being zero selects the output of the output multiplexor to be 7FFF FFFF. In the second saturated case, the saturation detection is based on the output of the zero detection circuit being zero (i.e., the minimum shift register does not contain all zeroes). In the third saturated case, the saturation detection is based on the MSB of the arithmetic shifter output,S31 184, and the MSB of the data,D31 182, being different, wherein the MSB of the data,D31 142, being one selects the output of the output multiplexor to be 8000 0000. In the fourth saturated case, the saturation detection is based on the output of the ones detection circuit being zero (i.e., the maximum shift register does not contain all ones). -
FIG. 2 illustrates a block diagram of a digitalsignal processing circuit 200 for implementing an arithmetic shifter and saturation detection circuit according to an embodiment of the present invention.FIG. 2 shows acontrol unit 210, adata unit 220 and anaddress unit 230. Thecontrol unit 210 directs the operation of the digital signal processor based on an instruction set (ISA) optimised for the task of rapid signal processing. The signal processing is divided between thecontrol unit 210 that directs program flow and one or more execution units that perform operations on data. Almost always, a collection of registers/memory 240 is included to hold operands and intermediate results. One of the execution units is theaddress unit 230. Theaddress unit 230, AU, directs the operand fetch for all variables which are defined and used by the executing instructions or program. Another execution unit is thedata unit 220, which includes at least onearithmetic logic unit 222 that accepts as inputs the data to be operated on and a code from thecontrol unit 210 indicating what operation to perform. Thearithmetic logic unit 222 provides the result of the computation and may indicate cases such as carry-in or carry-out, overflow, underflow and/or other statuses. -
FIG. 3 is aflow chart 300 of the method for providing a shift process with saturation for digital signal processor operations according to an embodiment of the present invention. First, data and shift amount values are provided to anarithmetic shifter 310. At the beginning of a cycle, the arithmetic shifter provides an output to maximum shifter and tominimum shifter 312. An underflow and overflow detection circuit detect underflow and overflow based on the content of the minimum shifter and themaximum shifter 314. A saturation decision multiplexor receives input from the underflow andoverflow detection circuits 316. A final decision multiplexor receives the output from the arithmetic shifter and the saturated value from theoutput multiplexor 318. Whether saturation occurs is determined by the final decision multiplexor using the output of the saturation decision multiplexor, the most significant bit of the arithmetic shifter output, and the most significant bit of data in 320. Saturation is detected if the most significant bit of the arithmetic shifter output and the most significant bit of data in are different or if the saturation decision multiplexor so indicates. When saturation is detected 322, the final decision multiplexor selects the saturate minimum or the saturate maximum depending on whether the MSB equals one or zero, respectively 330. Else 324, the final decision multiplexor selects the output from thearithmetic shifter 340. - A digital signal processor as illustrated above with respect to
FIGS. 1-3 may be used in a wide variety of products such as field programmable gate arrays, personal digital assistants (PDAs), cellular phones, digital scanners, digital cameras, personal computers and storage controllers.FIG. 4 shows one example wherein a magnetic disk drive includes a hard drive controller and microprocessor unit. Those skilled in the art will recognize that while the hard drive controller, microprocessor and memory, e.g., SRAM, may be shown separately, such components could easily be combined into one component, e.g., the hard drive controller. -
FIG. 4 shows a schematic block diagram showing a harddisk storage system 400 according to one embodiment of the present invention. The harddisk storage system 400 is connected to ahost computer 432. The harddisk storage system 400 responds to the write request by thehost computer 432 and records the recording data from thehost computer 432 on amagnetic disk 410, which serves as a recording medium. The harddisk storage system 400 further responds to the read request from thehost computer 432, reads the data recorded on themagnetic disk 410, and sends the data to thehost computer 432. The harddisk storage system 400 includes themagnetic disk 410, first andsecond motors head device 414, asignal processing circuit 420, aservo circuit 430, a microprocessor (MPU) 440, a memory (RAM) 450, a hard drive controller (HDC) 460, and aninterface circuit 470. The circuits 420-470 are connected to one another by abus 480. - The
magnetic disk 410 is rotated by thefirst motor 412 at a constant rotating speed. Thesecond motor 416 controls thehead device 414 so that it moves in the radial direction with respect to themagnetic disk 410. Thehead device 414 reads the data recorded on themagnetic disk 410 and sends a read signal, RD, to thesignal processing circuit 420. - The
signal processing circuit 420 samples the read signal, RD, in synchronism with a clock signal and generates a digital read signal. Thesignal processing circuit 420 carries out a decoding process on the digital read signal and outputs the decoded data signal. Theservo circuit 430 controls thefirst motor 412 and rotates themagnetic disk 410 at a constant speed. Theservo circuit 430 further receives the decoded data signal from thesignal processing circuit 420 via thebus 480 and controls thesecond motor 416 based on the servo data included in the digital read signal so that thehead device 414 is on track at the target position. - The
MPU 440 analyzes the write/read processing command sent from thehost computer 432 in accordance with the program data stored in theRAM 450 and sends a control signal to theHDC 460 via thebus 480. TheHDC 460 controls thesignal processing circuit 420 and theservo circuit 430 in accordance with the control signal from theMPU 440. TheHDC 460 further receives a data signal from thesignal processing circuit 420 via thebus 480. TheHDC 460 processes date, e.g., performs an error correcting code (ECC) process on the data signal. TheHDC 460 then sends the processed data to theinterface circuit 470 via thebus 480. Theinterface circuit 470 converts the data from theHDC 460 to a predetermined communication mode and sends the converted data to thehost computer 432. TheMPU 440 includes an arithmetic logic unit including an arithmetic shifter and saturation detection circuit as illustrated inFIG. 1 for shifting an operand to either maximum or the minimum value depending on the bit of data input when saturation occurs. According to one embodiment of the present invention, theHDC 460 sends a single shift instruction to theMPU 440. TheMPU 440 may be configured so that the shift instruction may be run by firmware of theHDC 460. In response to receiving the single shift instruction, theMPU 440 returns a shifted operand or a minimum/maximum saturation value. - It should be appreciated that the
MPU 440 could include a standalone processor or an embedded processor, e.g., theMPU 440 could be embedded in theHDC 460. TheMPU 440 could be part of a system on a chip (SOC). Further, theMPU 440 could be an ASIC, which would be hardware circuits that perform the function of the processor operating pursuant tomemory 450. In such a situation,memory 450 may be used but is not required, as the ASIC is designed to perform any assigned functions. It should also be appreciated thatmemory 450 could be either volatile or non-volatile memory. TheMPU 440 controls the operation of thevoice coil motor 416 andspindle motor 412 via theservo unit 430. - The foregoing description of the exemplary embodiment of the invention has been presented for the purposes of illustration and description. It is not intended to be exhaustive or to limit the invention to the precise form disclosed. Many modifications and variations are possible in light of the above teaching. It is intended that the scope of the invention be limited not with this detailed description, but rather by the claims appended hereto.
Claims (32)
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US11/068,206 US8209366B2 (en) | 2005-02-28 | 2005-02-28 | Method, apparatus and program storage device that provides a shift process with saturation for digital signal processor operations |
EP06000180A EP1696314A1 (en) | 2005-02-28 | 2006-01-05 | Arithmetic shifter, processor unit and storage device |
CNB2006100549550A CN100495316C (en) | 2005-02-28 | 2006-02-27 | Method, apparatus and program storage device that provides a shift process with saturation for digital signal processor operations |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US11/068,206 US8209366B2 (en) | 2005-02-28 | 2005-02-28 | Method, apparatus and program storage device that provides a shift process with saturation for digital signal processor operations |
Publications (2)
Publication Number | Publication Date |
---|---|
US20060195497A1 true US20060195497A1 (en) | 2006-08-31 |
US8209366B2 US8209366B2 (en) | 2012-06-26 |
Family
ID=36102994
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US11/068,206 Active 2027-12-08 US8209366B2 (en) | 2005-02-28 | 2005-02-28 | Method, apparatus and program storage device that provides a shift process with saturation for digital signal processor operations |
Country Status (3)
Country | Link |
---|---|
US (1) | US8209366B2 (en) |
EP (1) | EP1696314A1 (en) |
CN (1) | CN100495316C (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20090164544A1 (en) * | 2007-12-19 | 2009-06-25 | Jeffrey Dobbek | Dynamic range enhancement for arithmetic calculations in real-time control systems using fixed point hardware |
US20090182795A1 (en) * | 2007-12-19 | 2009-07-16 | Dobbek Jeffrey J | Method and system for performing calculations using fixed point microprocessor hardware |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9747105B2 (en) | 2009-12-17 | 2017-08-29 | Intel Corporation | Method and apparatus for performing a shift and exclusive or operation in a single instruction |
US9208839B2 (en) | 2014-03-20 | 2015-12-08 | Arm Limited | Predicting saturation in a shift operation |
CN112181354B (en) * | 2020-10-12 | 2021-08-10 | 上海芯旺微电子技术有限公司 | Method for synchronous shift saturation processing and application thereof |
CN112181355B (en) * | 2020-10-12 | 2021-08-06 | 上海芯旺微电子技术有限公司 | Shift saturation processing method and application thereof |
Citations (28)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US3812470A (en) * | 1972-07-31 | 1974-05-21 | Westinghouse Electric Corp | Programmable digital signal processor |
US4843543A (en) * | 1986-04-25 | 1989-06-27 | 501 Hitachi, Ltd. | Storage control method and apparatus |
US5241490A (en) * | 1992-01-06 | 1993-08-31 | Intel Corporation | Fully decoded multistage leading zero detector and normalization apparatus |
US5448509A (en) * | 1993-12-08 | 1995-09-05 | Hewlett-Packard Company | Efficient hardware handling of positive and negative overflow resulting from arithmetic operations |
US5557183A (en) * | 1993-07-29 | 1996-09-17 | International Business Machines Corporation | Method and apparatus for predicting failure of a disk drive |
US5684728A (en) * | 1994-06-10 | 1997-11-04 | Nec Corporation | Data processing system having a saturation arithmetic operation function |
US5706217A (en) * | 1996-07-30 | 1998-01-06 | Hughes Aircraft | Digital signal processing automatic gain control amplifier |
US5745393A (en) * | 1996-10-17 | 1998-04-28 | Samsung Electronics Company, Ltd. | Left-shifting an integer operand and providing a clamped integer result |
US5784532A (en) * | 1994-02-16 | 1998-07-21 | Qualcomm Incorporated | Application specific integrated circuit (ASIC) for performing rapid speech compression in a mobile telephone system |
US5844827A (en) * | 1996-10-17 | 1998-12-01 | Samsung Electronics Co., Ltd. | Arithmetic shifter that performs multiply/divide by two to the nth power for positive and negative N |
US5907498A (en) * | 1997-01-16 | 1999-05-25 | Samsung Electronics, Co., Ltd. | Circuit and method for overflow detection in a digital signal processor having a barrel shifter and arithmetic logic unit connected in series |
US5915109A (en) * | 1996-08-12 | 1999-06-22 | Mitsubishi Denki Kabushiki Kaisha | Microprocessor for processing a saturation instruction of an optional-bit length value |
US5959636A (en) * | 1996-02-23 | 1999-09-28 | Intel Corporation | Method and apparatus for performing saturation instructions using saturation limit values |
US5960200A (en) * | 1996-05-03 | 1999-09-28 | I-Cube | System to transition an enterprise to a distributed infrastructure |
US6078940A (en) * | 1997-01-24 | 2000-06-20 | Texas Instruments Incorporated | Microprocessor with an instruction for multiply and left shift with saturate |
US6314443B1 (en) * | 1998-11-20 | 2001-11-06 | Arm Limited | Double/saturate/add/saturate and double/saturate/subtract/saturate operations in a data processing system |
US6321248B1 (en) * | 1997-12-23 | 2001-11-20 | Stmicroelectronics S.A. | Process for determining an overflow to the format of the result of an arithmetic operation carried out on two operands |
US20020059351A1 (en) * | 2000-09-27 | 2002-05-16 | Mikio Shiraishi | Digital filter |
US6453464B1 (en) * | 1998-09-03 | 2002-09-17 | Legacyj. Corp., Inc. | Method and apparatus for converting COBOL to Java |
US6523171B1 (en) * | 1998-12-29 | 2003-02-18 | International Business Machines Corporation | Enhanced source code translator from procedural programming language (PPL) to an object oriented programming language (OOPL) |
US20030135813A1 (en) * | 2001-09-17 | 2003-07-17 | Mark Greenberg | Apparatus and method for saturating decoder values |
US6757819B1 (en) * | 2000-02-18 | 2004-06-29 | Texas Instruments Incorporated | Microprocessor with instructions for shifting data responsive to a signed count value |
US6822654B1 (en) * | 2001-12-31 | 2004-11-23 | Apple Computer, Inc. | Memory controller chipset |
US20060031272A1 (en) * | 2004-08-05 | 2006-02-09 | International Business Machines Corporation | Alignment shifter supporting multiple precisions |
US7103621B2 (en) * | 2002-03-29 | 2006-09-05 | Pts Corporation | Processor efficient transformation and lighting implementation for three dimensional graphics utilizing scaled conversion instructions |
US7206800B1 (en) * | 2000-08-30 | 2007-04-17 | Micron Technology, Inc. | Overflow detection and clamping with parallel operand processing for fixed-point multipliers |
US7213141B2 (en) * | 1999-12-14 | 2007-05-01 | Hitachi, Ltd. | Storage subsystem and storage controller |
US7543014B2 (en) * | 2002-07-31 | 2009-06-02 | Texas Instruments Incorporated | Saturated arithmetic in a processing unit |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1058185A1 (en) | 1999-05-31 | 2000-12-06 | Motorola, Inc. | A multiply and accumulate apparatus and a method thereof |
-
2005
- 2005-02-28 US US11/068,206 patent/US8209366B2/en active Active
-
2006
- 2006-01-05 EP EP06000180A patent/EP1696314A1/en not_active Withdrawn
- 2006-02-27 CN CNB2006100549550A patent/CN100495316C/en not_active Expired - Fee Related
Patent Citations (28)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US3812470A (en) * | 1972-07-31 | 1974-05-21 | Westinghouse Electric Corp | Programmable digital signal processor |
US4843543A (en) * | 1986-04-25 | 1989-06-27 | 501 Hitachi, Ltd. | Storage control method and apparatus |
US5241490A (en) * | 1992-01-06 | 1993-08-31 | Intel Corporation | Fully decoded multistage leading zero detector and normalization apparatus |
US5557183A (en) * | 1993-07-29 | 1996-09-17 | International Business Machines Corporation | Method and apparatus for predicting failure of a disk drive |
US5448509A (en) * | 1993-12-08 | 1995-09-05 | Hewlett-Packard Company | Efficient hardware handling of positive and negative overflow resulting from arithmetic operations |
US5784532A (en) * | 1994-02-16 | 1998-07-21 | Qualcomm Incorporated | Application specific integrated circuit (ASIC) for performing rapid speech compression in a mobile telephone system |
US5684728A (en) * | 1994-06-10 | 1997-11-04 | Nec Corporation | Data processing system having a saturation arithmetic operation function |
US5959636A (en) * | 1996-02-23 | 1999-09-28 | Intel Corporation | Method and apparatus for performing saturation instructions using saturation limit values |
US5960200A (en) * | 1996-05-03 | 1999-09-28 | I-Cube | System to transition an enterprise to a distributed infrastructure |
US5706217A (en) * | 1996-07-30 | 1998-01-06 | Hughes Aircraft | Digital signal processing automatic gain control amplifier |
US5915109A (en) * | 1996-08-12 | 1999-06-22 | Mitsubishi Denki Kabushiki Kaisha | Microprocessor for processing a saturation instruction of an optional-bit length value |
US5844827A (en) * | 1996-10-17 | 1998-12-01 | Samsung Electronics Co., Ltd. | Arithmetic shifter that performs multiply/divide by two to the nth power for positive and negative N |
US5745393A (en) * | 1996-10-17 | 1998-04-28 | Samsung Electronics Company, Ltd. | Left-shifting an integer operand and providing a clamped integer result |
US5907498A (en) * | 1997-01-16 | 1999-05-25 | Samsung Electronics, Co., Ltd. | Circuit and method for overflow detection in a digital signal processor having a barrel shifter and arithmetic logic unit connected in series |
US6078940A (en) * | 1997-01-24 | 2000-06-20 | Texas Instruments Incorporated | Microprocessor with an instruction for multiply and left shift with saturate |
US6321248B1 (en) * | 1997-12-23 | 2001-11-20 | Stmicroelectronics S.A. | Process for determining an overflow to the format of the result of an arithmetic operation carried out on two operands |
US6453464B1 (en) * | 1998-09-03 | 2002-09-17 | Legacyj. Corp., Inc. | Method and apparatus for converting COBOL to Java |
US6314443B1 (en) * | 1998-11-20 | 2001-11-06 | Arm Limited | Double/saturate/add/saturate and double/saturate/subtract/saturate operations in a data processing system |
US6523171B1 (en) * | 1998-12-29 | 2003-02-18 | International Business Machines Corporation | Enhanced source code translator from procedural programming language (PPL) to an object oriented programming language (OOPL) |
US7213141B2 (en) * | 1999-12-14 | 2007-05-01 | Hitachi, Ltd. | Storage subsystem and storage controller |
US6757819B1 (en) * | 2000-02-18 | 2004-06-29 | Texas Instruments Incorporated | Microprocessor with instructions for shifting data responsive to a signed count value |
US7206800B1 (en) * | 2000-08-30 | 2007-04-17 | Micron Technology, Inc. | Overflow detection and clamping with parallel operand processing for fixed-point multipliers |
US20020059351A1 (en) * | 2000-09-27 | 2002-05-16 | Mikio Shiraishi | Digital filter |
US20030135813A1 (en) * | 2001-09-17 | 2003-07-17 | Mark Greenberg | Apparatus and method for saturating decoder values |
US6822654B1 (en) * | 2001-12-31 | 2004-11-23 | Apple Computer, Inc. | Memory controller chipset |
US7103621B2 (en) * | 2002-03-29 | 2006-09-05 | Pts Corporation | Processor efficient transformation and lighting implementation for three dimensional graphics utilizing scaled conversion instructions |
US7543014B2 (en) * | 2002-07-31 | 2009-06-02 | Texas Instruments Incorporated | Saturated arithmetic in a processing unit |
US20060031272A1 (en) * | 2004-08-05 | 2006-02-09 | International Business Machines Corporation | Alignment shifter supporting multiple precisions |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20090164544A1 (en) * | 2007-12-19 | 2009-06-25 | Jeffrey Dobbek | Dynamic range enhancement for arithmetic calculations in real-time control systems using fixed point hardware |
US20090182795A1 (en) * | 2007-12-19 | 2009-07-16 | Dobbek Jeffrey J | Method and system for performing calculations using fixed point microprocessor hardware |
US8280941B2 (en) | 2007-12-19 | 2012-10-02 | HGST Netherlands B.V. | Method and system for performing calculations using fixed point microprocessor hardware |
Also Published As
Publication number | Publication date |
---|---|
EP1696314A1 (en) | 2006-08-30 |
CN100495316C (en) | 2009-06-03 |
CN1828520A (en) | 2006-09-06 |
US8209366B2 (en) | 2012-06-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110036368B (en) | Apparatus and method for performing arithmetic operations to accumulate floating point numbers | |
JP6001276B2 (en) | Apparatus and method for performing floating point addition | |
US8209366B2 (en) | Method, apparatus and program storage device that provides a shift process with saturation for digital signal processor operations | |
EP1058185A1 (en) | A multiply and accumulate apparatus and a method thereof | |
US7716268B2 (en) | Method and apparatus for providing a processor based nested form polynomial engine | |
US9870200B2 (en) | Decimal and binary floating point rounding | |
CN108139912B (en) | Apparatus and method for calculating and preserving error bounds during floating point operations | |
US7668892B2 (en) | Data processing apparatus and method for normalizing a data value | |
US4992969A (en) | Integer division circuit provided with a overflow detector circuit | |
MX2008014047A (en) | Controlled-precision iterative arithmetic logic unit. | |
US20050223055A1 (en) | Method and apparatus to correct leading one prediction | |
EP0350928B1 (en) | Data processor capable of executing division of signed data with a small number of program steps | |
JP4402654B2 (en) | Result segmentation in SIMD data processing system | |
JPH04350724A (en) | Shift extent detecting circuit | |
US6393452B1 (en) | Method and apparatus for performing load bypasses in a floating-point unit | |
CN111290790A (en) | Conversion device for converting fixed point into floating point | |
US6760738B2 (en) | Exponent unit of data processing system | |
JP2591250B2 (en) | Data processing device | |
JPS58186840A (en) | Data processor | |
JP2001092634A (en) | Functional unit and computing method and processor | |
JPH1031618A (en) | Data conversion system | |
JPH03245226A (en) | Floating decimal-point multiplier | |
JPS5851352A (en) | Data processor | |
JPH09218785A (en) | Information processor |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: HITACHI GLOBAL STORAGE TECHNOLOGIES NETHERLANDS B. Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:DOBBEK, JEFFREY J.;HWANG, KIRK;REEL/FRAME:016010/0144 Effective date: 20050215 |
|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |
|
AS | Assignment |
Owner name: HGST, NETHERLANDS B.V., NETHERLANDS Free format text: CHANGE OF NAME;ASSIGNOR:HGST, NETHERLANDS B.V.;REEL/FRAME:029341/0777 Effective date: 20120723 Owner name: HGST NETHERLANDS B.V., NETHERLANDS Free format text: CHANGE OF NAME;ASSIGNOR:HITACHI GLOBAL STORAGE TECHNOLOGIES NETHERLANDS B.V.;REEL/FRAME:029341/0777 Effective date: 20120723 |
|
FPAY | Fee payment |
Year of fee payment: 4 |
|
AS | Assignment |
Owner name: WESTERN DIGITAL TECHNOLOGIES, INC., CALIFORNIA Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:HGST NETHERLANDS B.V.;REEL/FRAME:040819/0450 Effective date: 20160831 |
|
MAFP | Maintenance fee payment |
Free format text: PAYMENT OF MAINTENANCE FEE, 8TH YEAR, LARGE ENTITY (ORIGINAL EVENT CODE: M1552); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY Year of fee payment: 8 |
|
AS | Assignment |
Owner name: JPMORGAN CHASE BANK, N.A., AS AGENT, ILLINOIS Free format text: SECURITY INTEREST;ASSIGNOR:WESTERN DIGITAL TECHNOLOGIES, INC.;REEL/FRAME:052915/0566 Effective date: 20200113 |
|
AS | Assignment |
Owner name: WESTERN DIGITAL TECHNOLOGIES, INC., CALIFORNIA Free format text: RELEASE OF SECURITY INTEREST AT REEL 052915 FRAME 0566;ASSIGNOR:JPMORGAN CHASE BANK, N.A.;REEL/FRAME:059127/0001 Effective date: 20220203 |
|
AS | Assignment |
Owner name: JPMORGAN CHASE BANK, N.A., ILLINOIS Free format text: PATENT COLLATERAL AGREEMENT - A&R LOAN AGREEMENT;ASSIGNOR:WESTERN DIGITAL TECHNOLOGIES, INC.;REEL/FRAME:064715/0001 Effective date: 20230818 |
|
MAFP | Maintenance fee payment |
Free format text: PAYMENT OF MAINTENANCE FEE, 12TH YEAR, LARGE ENTITY (ORIGINAL EVENT CODE: M1553); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY Year of fee payment: 12 |